CN103457799B - Microblog zombie user detection method based on graph of a relation - Google Patents
Microblog zombie user detection method based on graph of a relation Download PDFInfo
- Publication number
- CN103457799B CN103457799B CN201310396404.2A CN201310396404A CN103457799B CN 103457799 B CN103457799 B CN 103457799B CN 201310396404 A CN201310396404 A CN 201310396404A CN 103457799 B CN103457799 B CN 103457799B
- Authority
- CN
- China
- Prior art keywords
- user
- users
- relation
- sample
- graph
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The microblog zombie user detection method analyzed based on graph of a relation, analyzes module including data collection module and graph of a relation；Data collection module is for collecting the data of a known corpse user and selecting sample of users from data.Whether graph of a relation analyzes module for judgment sample user is corpse user: initially set up known corpse user and the graph of a relation of sample of users；The malice scoring of initialising subscriber；Calculate the relatedness numerical value of user and calculate and update the malice scoring of sample of users according to graph of a relation and propagation rule；The whether propagation convergence of the malice scoring of judgment sample user；Judge that the malice scoring propagating the sample of users of convergence, whether more than threshold value, is corpse user more than threshold value then this sample of users.The present invention utilizes the social relations of corpse user and semantic relation to find and identify other corpses user, improves the identification effect of corpse user.The present invention applies to social networks, more safely and effectively detects service for its offer, improves the safety of social networks.
Description
Technical field
The present invention relates to a kind of microblog zombie user detection method, particularly relate to a kind of microblogging corpse based on graph of a relation
User's detection method.
Background technology
Current Technology Times, popular along with intelligent communications terminal, mobile Internet comes into people's the most day by day
In daily life.Currently, social networks is application quite popular in mobile Internet, as external
" Facebook " and " Twitter " etc., in China, online social networks becomes one mainly the most gradually
Platform, compare know and conventional be microblogging, people collect information by microblogging on network, make friends with will
The people that people having a common goal closes.Microblogging, the i.e. abbreviation of micro-blog (MicroBlog), be that one can give out information immediately
The system of similar blog, be an Information Sharing based on customer relationship, propagate and obtain platform.Micro-
Rich " Twitter " being referred to as China, because it is similar to Twitter.Microblogging is possible not only to meet user's phase
Recommend the network consulting interested mutually, it is also possible to pay close attention to famous person or friend that oneself is appreciated, check others
Latest tendency or deliver oneself up-to-date speech, thus reach to share happy purpose with others, the most micro-
Win liking of more and more people.
But, from the point of view of the use habit and different angle of culturals of user, microblogging and " Twitter " they are not
With.According to the research of HP Lab, people prefer forwarding information on microblogging rather than deliver original
Microblogging, as long as that shares is valuable, people on microblogging can very kind help forward.Additionally, both
Consumer's Experience aspect is the most different.At " Twitter ", people can only share text message, but in microblogging,
People can also share picture, video and audio frequency.Additionally, microblogging allows also to user's recoil state and same
Time forwards, and this is infeasible at " Twitter ".
Along with the broad development of microblogging, occur in that many fictitious users, i.e. corpse user.These corpses user
Appearance exist the most many reasons.On the one hand, in order to meet the vanity of microblog users and improve individual
The attention rate of microblogging, some people selects to spend money on the user of some falsenesses to improve the user's silk number of oneself,
This behavior greatly reduces the personal integrity of user；Another aspect is exactly this detection that people utilizes microblogging
Leak peddles " corpse user " without restraint, and those people controlling " corpse user " rearward are obtained by transaction
No small interests, cause the generation of corpse user's industrial chain, bring no small negative effect for microblogging.This is also
Being microblogging is clearly distinguished from one of " Twitter ".
Now, many people are had to study west social network sites, but, the research to China's social networks is but one
Sheet is blank.Due to the greatest differences of microblogging He " Twitter ", so, those skilled in the art is devoted to out
The method sending out a kind of to detect microblogging corpse user.
Summary of the invention
Because the drawbacks described above of prior art, the technical problem to be solved is to provide a kind of based on visually
Change the microblog zombie user detection method that graph of a relation is analyzed.
For achieving the above object, the invention provides a kind of microblogging corpse user side of detection based on visible relation network
Method, it is characterised in that include that data collection module and graph of a relation analyze module；
Described data collection module, for according to a known corpse user, collects the number of described known corpse user
According to, and pick out sample of users；
Described graph of a relation is analyzed module and is used for judging whether described sample of users is corpse user, specifically includes following step
Rapid:
Step 201, visualizes the attribute of a relation of described known corpse user and described sample of users, makes
Graph of a relation: described known corpse user and described sample of users are all as the node of described graph of a relation；
Step 202, initializes the malice scoring of described known corpse user and described sample of users；
Step 203, analyzes the general character of described graph of a relation, calculates the pass of each described node in described graph of a relation
Connection property numerical value, and calculate according to propagation rule and described graph of a relation and update the malice of described sample of users and comment
Point；
Step 204, it is judged that whether the described malice scoring of described sample of users propagates convergence, receives if propagated
Hold back, jump into step 205；If not propagating convergence, then redirect into step 203；
Step 205, it is judged that whether the described malice scoring of described sample of users is more than threshold value, if greater than described threshold
Value, then redirect into step 206；If less than described threshold value, the most described sample of users is judged as normal users；
Step 206: described sample of users is judged as corpse user；
Step 207: be disposed.
Further, described data collection module is the number of the described known corpse user collected by microblogging API
According to.
Further, the data of described known corpse user include user's vermicelli and the name of follower and quantity.
Further, described data collection module is random to the selection of described sample of users.
Further, described data collection module selects user's vermicelli of described sample of users and the quantity of follower
Less than 1000.
Further, there is the pass paid close attention to be concerned between the adjacent node of the described graph of a relation of described step 201
System.
Further, in described step 202, the malice scoring of described known corpse user is initialized as 1, institute
The malice scoring stating sample of users is initialized as 0.
Further, in described step 203, the described relatedness numerical value of described node is corresponding to described node
The inverse of vermicelli quantity of user.
Further, the described propagation rule in described step 203 includes:
A), when calculating the scoring of the malice of vermicelli of a user, the malice scoring that malice scoring is user of vermicelli
It is multiplied by the relatedness numerical value of user；
B), when a user pays close attention to multiple user, the malice scoring of a user is multiple users that it is paid close attention to
Malice scoring sum.
Further, the described propagation convergence of described step 204 refers to that the described malice scoring of described sample of users reaches
No longer change to stable.
Owing to microblogging lacks the testing mechanism to corpse user, by dividing corpse customer relationship network comprehensively
Analysis, the detection method accuracy being trained out and recurrence degree are the highest, and combination property is preferable.In being applicable to
The corpse user of state's social networks differentiates.Owing to corpse user's great majority are to be automatically generated by system, so he
ID be largely similar, in addition in order to avoid being detected, corpse user is the most mutual
Concern makes to look as broad as long with normal users, therefore has many ID in their social network diagram
Similar user gathers at one piece, if so a corpse user can be found in advance the most probably to look for
To relative corpse user, it is greatly improved the efficiency of reasoning algorithm.
Below with reference to accompanying drawing, the technique effect of design, concrete structure and the generation of the present invention is described further,
To be fully understood from the purpose of the present invention, feature and effect.
Accompanying drawing explanation
Fig. 1 is the processing procedure of the data collection module of the present invention；
Fig. 2 is the process chart of the graph of a relation analysis module of the present invention.
Detailed description of the invention
Below in conjunction with the accompanying drawings embodiments of the invention are elaborated: the present embodiment is with the technology of the present invention side
Implement under case premise, give detailed embodiment and concrete operating process, but the guarantor of the present invention
The scope of protecting is not limited to following embodiment.
A kind of based on graph of a relation the microblog zombie user detection method of the present invention, is divided into two big modules: data collection
Module and graph of a relation analyze module.
The handling process of data collection module is concrete as shown in Figure 1.First, by the API(Application of microblogging
Programming Interface, application programming interface) 101 from the beginning of the account of known corpse user, receive
Collect user's vermicelli of the data 102 of known corpse user, i.e. this corpse user and the user name of follower and quantity；
Then, user's vermicelli and the data of follower of known corpse user are collected；Finally, from the use of known corpse user
Family vermicelli and follower select sample of users, and mixes the sample with family and known corpse user data is stored in relation data
In storehouse, wherein sample of users randomly chooses.In the preferred embodiment, in order to ensure select with
Machine, in user's vermicelli and follower of known corpse user, selects its user's vermicelli and follower to be less than
The user of 1000 is as sample of users.
The process of data collection module is the user profile obtained by manual entry microblogging: each microblog users
Having the user name of oneself, and according to user name, each user has a link of the personal page:
Http:// weibo.com/userid, signs in in this page, the data (user's vermicelli and follower) of user
Can open-and-shut find.
Graph of a relation analyzes the handling process of module as in figure 2 it is shown, specifically include following steps:
Step 201, visualizes the attribute of a relation of known corpse user and sample of users, makes graph of a relation:
Each user (including that known corpse includes known corpse user and sample of users) is regarded as a node,
For any two user, if having the relation paid close attention to be concerned between them, then between the two node
Being connected by a directed line segment, the user being concerned is pointed in direction by vermicelli.
Step 202, initializes the malice of each node users (including known corpse user and sample of users)
Scoring: the malice scoring setting known corpse user is 1, the malice scoring setting each sample of users is 0；
Step 203, analyzes the general character of graph of a relation and draws the social relations of corpse user, calculate and update sample
The malice scoring of this user:
1) the relatedness numerical value of user is calculated: add up the vermicelli quantity of each user, and by this quantity
The inverted relatedness numerical value being between this user and its vermicelli；
2) this relatedness numerical value is adjacent as this user the weights on the limit of the graph of a relation of user；
3) press propagation rule according to the malice scoring of relatedness numerical value and known corpse user and calculate sample of users
Malice scoring, its propagation rule is: a) when calculate a user vermicelli malice scoring time, vermicelli
Malice scoring is that the relatedness numerical value of user is multiplied by the malice scoring of user；B) pay close attention to multiple as a user
During user, the scoring of this user will be the malice scoring sum of all users that this user is paid close attention to；
4) according to graph of a relation iterative computation, the malice scoring of each sample of users is updated.
Step 204, it is judged that whether the malice scoring of sample of users reaches stationary value no longer changes, and is propagation
Convergence: if propagating convergence, then jump to step 205；Without reaching to propagate convergence, then jump to
Step 203.
Step 205, judges each sample of users malice score value now: if malice scoring is more than threshold
Value (threshold value determines and can be determined by enlightenment experiment), then redirect such as step 206；If malice scoring is less than threshold value,
Then this user is judged as normal users.
Step 206: this sample of users is judged as corpse user.
Step 207: corpse user judges to terminate.
The preferred embodiment of the present invention described in detail above.Should be appreciated that the ordinary skill of this area without
Creative work just can make many modifications and variations according to the design of the present invention.Therefore, all in the art
Technical staff passes through logical analysis, reasoning, or a limited experiment the most on the basis of existing technology
Available technical scheme, all should be in the protection domain being defined in the patent claims.
Claims (10)
1. a microblog zombie user detection method based on visible relation network, it is characterised in that include that data are received
Collection module and graph of a relation analyze module；
Described data collection module, for according to a known corpse user, collects the number of described known corpse user
According to, and pick out sample of users；
Described graph of a relation is analyzed module and is used for judging whether described sample of users is corpse user, specifically includes following
Step:
Step (201), visualizes the attribute of a relation of described known corpse user and described sample of users,
Make graph of a relation: described known corpse user and described sample of users are all as the joint of described graph of a relation
Point；
Step (202), initializes the malice scoring of described known corpse user and described sample of users；
Step (203), analyzes the general character of described graph of a relation, calculates each described node in described graph of a relation
Relatedness numerical value, and calculate according to propagation rule and described graph of a relation and update described sample of users
Malice scoring；
Step (204), it is judged that whether the described malice scoring of described sample of users propagates convergence, if passed
Broadcast convergence, jump into step (205)；If not propagating convergence, then redirect into step (203)；
Step (205), it is judged that whether the described malice scoring of described sample of users is more than threshold value, if greater than institute
State threshold value, then redirect into step (206)；If less than described threshold value, the most described sample of users is determined
For normal users；
Step (206), described sample of users is judged as corpse user；
Step (207), is disposed.
2. microblog zombie user detection method as claimed in claim 1, wherein, described data collection module is logical
Cross microblogging API and collect the data of described known corpse user.
3. microblog zombie user detection method as claimed in claim 1, wherein, the number of described known corpse user
According to the name and the quantity that include user's vermicelli and follower.
4. microblog zombie user detection method as claimed in claim 1, wherein, described data collection module is to institute
The selection stating sample of users is random.
5. microblog zombie user detection method as claimed in claim 1, wherein, described data collection module selects
User's vermicelli of described sample of users and the quantity of follower less than 1000.
6. microblog zombie user detection method as claimed in claim 1, wherein, the institute of described step (201)
State and between the adjacent node of graph of a relation, have the relation paid close attention to be concerned.
7. microblog zombie user detection method as claimed in claim 1, wherein, in described step (202),
The malice scoring of described known corpse user is initialized as 1, and the malice scoring of described sample of users initializes
It is 0.
8. microblog zombie user detection method as claimed in claim 1, wherein, in described step (203),
The described relatedness numerical value of described node is the inverse of the vermicelli quantity of the user corresponding to described node.
9. microblog zombie user detection method as claimed in claim 1, wherein, institute in described step (203)
State propagation rule to include:
A), when calculating the scoring of the malice of vermicelli of a user, the malice that malice scoring is user of vermicelli is commented
Take the relatedness numerical value with user separately；
B), when a user pays close attention to multiple user, the malice scoring of a user is multiple use that it is paid close attention to
The malice scoring sum at family.
10. microblog zombie user detection method as claimed in claim 1, wherein, institute in described step (204)
The convergence of propagating stated refers to that the described malice scoring of described sample of users reaches stable and no longer changes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310396404.2A CN103457799B (en) | 2013-09-03 | 2013-09-03 | Microblog zombie user detection method based on graph of a relation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310396404.2A CN103457799B (en) | 2013-09-03 | 2013-09-03 | Microblog zombie user detection method based on graph of a relation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103457799A CN103457799A (en) | 2013-12-18 |
| CN103457799B true CN103457799B (en) | 2016-08-17 |
Family
ID=49739784
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310396404.2A Expired - Fee Related CN103457799B (en) | 2013-09-03 | 2013-09-03 | Microblog zombie user detection method based on graph of a relation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103457799B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104092601B (en) * | 2014-07-28 | 2017-12-05 | 北京微众文化传媒有限公司 | The recognition methods of social networks account and device |
| CN105005594B (en) * | 2015-06-29 | 2018-07-13 | 嘉兴慧康智能科技有限公司 | Abnormal microblog users recognition methods |
| CN105357189B (en) * | 2015-10-13 | 2018-05-01 | 精硕科技（北京）股份有限公司 | Corpse account detection method and device |
| CN105389505B (en) * | 2015-10-19 | 2018-06-12 | 西安电子科技大学 | Support attack detection method based on the sparse self-encoding encoder of stack |
| CN105956184B (en) * | 2016-06-01 | 2017-05-31 | 西安交通大学 | Collaborative and organized junk information issue the recognition methods of group in a kind of microblogging community network |
| CN108471445B (en) * | 2018-04-02 | 2021-08-13 | 北京奇艺世纪科技有限公司 | Attention object content updating notification method and device |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102946331A (en) * | 2012-10-10 | 2013-02-27 | 北京交通大学 | Detecting method and device for zombie users of social networks |
| CN103117891A (en) * | 2013-01-18 | 2013-05-22 | 武汉大学 | Zombie user detection method on microblog platform |
-
2013
- 2013-09-03 CN CN201310396404.2A patent/CN103457799B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102946331A (en) * | 2012-10-10 | 2013-02-27 | 北京交通大学 | Detecting method and device for zombie users of social networks |
| CN103117891A (en) * | 2013-01-18 | 2013-05-22 | 武汉大学 | Zombie user detection method on microblog platform |
Non-Patent Citations (3)
| Title |
|---|
| Detecting Automation of Twitter Accounts: Are You a Human, Bot, or Cyborg;Zi Chu, Steven Gianvecchio, Haining Wang, Sushil Jajodia;《IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING》;20121231;第9卷;全文 * |
| 一种新型智能僵尸粉甄别方法;方明，方意;《计算机工程》;20130430;全文 * |
| 一种降低微博僵尸粉影响的方法;原福永，冯静，符茜茜，曹旭峰;《现代图书情报技术》;20120531;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103457799A (en) | 2013-12-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103457799B (en) | Microblog zombie user detection method based on graph of a relation | |
| Adali et al. | Measuring behavioral trust in social networks | |
| US9542503B2 (en) | Estimation of closeness of topics based on graph analytics | |
| CN104519107B (en) | A kind of method and device for expanding contact person | |
| US10074094B2 (en) | Generating a user profile based on self disclosed public status information | |
| US8838564B2 (en) | Method to increase content relevance using insights obtained from user activity updates | |
| US8600901B2 (en) | Providing social likeness within a messaging context | |
| Lam et al. | Involuntary information leakage in social network services | |
| US20090319288A1 (en) | Suggesting contacts for social networks | |
| US20100076775A1 (en) | Progressive capture of prospect information for user profiles | |
| KR102009912B1 (en) | Method and system to automatically change or update the configuration or setting of a communication system | |
| KR101313791B1 (en) | Method for providing topic based social network service | |
| KR20100130003A (en) | Apparatus and method for providing online social network service | |
| US20100070372A1 (en) | Using spam and user actions to infer advertisements | |
| US9613339B2 (en) | Information exchange in the social network environment | |
| CN104468321B (en) | The method and system of social interaction is initiated based on Email content | |
| CN105550275A (en) | Microblog forwarding quantity prediction method | |
| CN103425703A (en) | Method and device for processing network information | |
| CN103279483B (en) | A kind of topic Epidemic Scope appraisal procedure towards micro-blog and system | |
| CN102546462A (en) | Method of recommending contact persons in instant messaging, client side and system | |
| CN103200073B (en) | A kind of information processing method and device | |
| Xu et al. | Using physical context in a mobile social networking application for improving friend recommendations | |
| CN106603387B (en) | Method and system for predicting microblog forwarding path based on microblog forwarding relation | |
| Jeong et al. | The social relation key: A new paradigm for security | |
| CN102750288A (en) | Internet content recommending method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20160817 Termination date: 20210903 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |